| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.498407643312102, | |
| "eval_steps": 500, | |
| "global_step": 1569, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01592356687898089, | |
| "grad_norm": 4.3297224044799805, | |
| "learning_rate": 2.5316455696202533e-05, | |
| "loss": 0.86, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03184713375796178, | |
| "grad_norm": 6.637640476226807, | |
| "learning_rate": 5.0632911392405066e-05, | |
| "loss": 0.4332, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04777070063694268, | |
| "grad_norm": 1.9116368293762207, | |
| "learning_rate": 7.59493670886076e-05, | |
| "loss": 0.3014, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06369426751592357, | |
| "grad_norm": 1.2928426265716553, | |
| "learning_rate": 0.00010126582278481013, | |
| "loss": 0.2232, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07961783439490445, | |
| "grad_norm": 1.4354435205459595, | |
| "learning_rate": 0.00012658227848101267, | |
| "loss": 0.1771, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09554140127388536, | |
| "grad_norm": 1.710893154144287, | |
| "learning_rate": 0.0001518987341772152, | |
| "loss": 0.2136, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11146496815286625, | |
| "grad_norm": 1.6920921802520752, | |
| "learning_rate": 0.00017721518987341773, | |
| "loss": 0.1644, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 1.2916326522827148, | |
| "learning_rate": 0.00019999977772170748, | |
| "loss": 0.1593, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14331210191082802, | |
| "grad_norm": 1.599898099899292, | |
| "learning_rate": 0.00019997310552224846, | |
| "loss": 0.1578, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 1.4364933967590332, | |
| "learning_rate": 0.00019990199125039174, | |
| "loss": 0.155, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1751592356687898, | |
| "grad_norm": 1.7148686647415161, | |
| "learning_rate": 0.00019978646651929572, | |
| "loss": 0.1463, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 1.2754517793655396, | |
| "learning_rate": 0.0001996265826843568, | |
| "loss": 0.1471, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2070063694267516, | |
| "grad_norm": 1.0604585409164429, | |
| "learning_rate": 0.00019942241082037982, | |
| "loss": 0.124, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2229299363057325, | |
| "grad_norm": 0.6741822361946106, | |
| "learning_rate": 0.00019917404168998256, | |
| "loss": 0.0979, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23885350318471338, | |
| "grad_norm": 1.2134864330291748, | |
| "learning_rate": 0.00019888158570324795, | |
| "loss": 0.1105, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 0.6342090368270874, | |
| "learning_rate": 0.00019854517286864245, | |
| "loss": 0.099, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27070063694267515, | |
| "grad_norm": 1.0112156867980957, | |
| "learning_rate": 0.000198164952735222, | |
| "loss": 0.0928, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.28662420382165604, | |
| "grad_norm": 1.0317986011505127, | |
| "learning_rate": 0.00019774109432615147, | |
| "loss": 0.0963, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.30254777070063693, | |
| "grad_norm": 0.7419966459274292, | |
| "learning_rate": 0.00019727378606356703, | |
| "loss": 0.1069, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 1.373571753501892, | |
| "learning_rate": 0.00019676323568481498, | |
| "loss": 0.1103, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3343949044585987, | |
| "grad_norm": 0.6964902877807617, | |
| "learning_rate": 0.00019620967015010395, | |
| "loss": 0.0827, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3503184713375796, | |
| "grad_norm": 0.9365194439888, | |
| "learning_rate": 0.00019561333554161224, | |
| "loss": 0.1046, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3662420382165605, | |
| "grad_norm": 1.899908423423767, | |
| "learning_rate": 0.00019497449695409408, | |
| "loss": 0.1031, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 0.8126282095909119, | |
| "learning_rate": 0.00019429343837703455, | |
| "loss": 0.1069, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3980891719745223, | |
| "grad_norm": 1.244973063468933, | |
| "learning_rate": 0.00019357046256840473, | |
| "loss": 0.082, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4140127388535032, | |
| "grad_norm": 0.8132479786872864, | |
| "learning_rate": 0.00019280589092007352, | |
| "loss": 0.0867, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4299363057324841, | |
| "grad_norm": 0.8742226958274841, | |
| "learning_rate": 0.0001920000633149362, | |
| "loss": 0.0761, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.445859872611465, | |
| "grad_norm": 0.6832584738731384, | |
| "learning_rate": 0.00019115333797582254, | |
| "loss": 0.0845, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.46178343949044587, | |
| "grad_norm": 0.6030533909797668, | |
| "learning_rate": 0.00019026609130625257, | |
| "loss": 0.0722, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 0.7257299423217773, | |
| "learning_rate": 0.0001893387177231099, | |
| "loss": 0.0701, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.49363057324840764, | |
| "grad_norm": 0.6390131711959839, | |
| "learning_rate": 0.00018837162948130752, | |
| "loss": 0.0696, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 1.8211143016815186, | |
| "learning_rate": 0.00018736525649052394, | |
| "loss": 0.0757, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5254777070063694, | |
| "grad_norm": 1.0733026266098022, | |
| "learning_rate": 0.00018632004612409103, | |
| "loss": 0.0835, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5414012738853503, | |
| "grad_norm": 0.7683069705963135, | |
| "learning_rate": 0.00018523646302011867, | |
| "loss": 0.0765, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5573248407643312, | |
| "grad_norm": 0.5842871069908142, | |
| "learning_rate": 0.00018411498887494396, | |
| "loss": 0.073, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 0.6933338046073914, | |
| "learning_rate": 0.0001829561222289984, | |
| "loss": 0.0648, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.589171974522293, | |
| "grad_norm": 0.42543935775756836, | |
| "learning_rate": 0.000181760378245186, | |
| "loss": 0.072, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6050955414012739, | |
| "grad_norm": 0.72137451171875, | |
| "learning_rate": 0.0001805282884798732, | |
| "loss": 0.0661, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6210191082802548, | |
| "grad_norm": 0.6393111944198608, | |
| "learning_rate": 0.00017926040064659014, | |
| "loss": 0.0841, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 0.46522364020347595, | |
| "learning_rate": 0.00017795727837255015, | |
| "loss": 0.0668, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6528662420382165, | |
| "grad_norm": 0.790262758731842, | |
| "learning_rate": 0.0001766195009480949, | |
| "loss": 0.0702, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6687898089171974, | |
| "grad_norm": 0.7343723177909851, | |
| "learning_rate": 0.00017524766306917618, | |
| "loss": 0.0751, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6847133757961783, | |
| "grad_norm": 0.497380793094635, | |
| "learning_rate": 0.00017384237457298987, | |
| "loss": 0.0663, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7006369426751592, | |
| "grad_norm": 0.5845641493797302, | |
| "learning_rate": 0.00017240426016687863, | |
| "loss": 0.0656, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7165605095541401, | |
| "grad_norm": 1.0296058654785156, | |
| "learning_rate": 0.00017093395915062428, | |
| "loss": 0.0664, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.732484076433121, | |
| "grad_norm": 0.7704136371612549, | |
| "learning_rate": 0.00016943212513225345, | |
| "loss": 0.0597, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7484076433121019, | |
| "grad_norm": 0.7734206318855286, | |
| "learning_rate": 0.00016789942573748232, | |
| "loss": 0.0638, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 0.6289462447166443, | |
| "learning_rate": 0.00016633654231293013, | |
| "loss": 0.0573, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7802547770700637, | |
| "grad_norm": 0.6616397500038147, | |
| "learning_rate": 0.00016474416962323325, | |
| "loss": 0.0506, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 0.3910081386566162, | |
| "learning_rate": 0.00016312301554219426, | |
| "loss": 0.0457, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8121019108280255, | |
| "grad_norm": 0.6165742874145508, | |
| "learning_rate": 0.00016147380073810346, | |
| "loss": 0.0503, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8280254777070064, | |
| "grad_norm": 0.6931295990943909, | |
| "learning_rate": 0.00015979725835337294, | |
| "loss": 0.0481, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8439490445859873, | |
| "grad_norm": 0.7499469518661499, | |
| "learning_rate": 0.00015809413367862512, | |
| "loss": 0.0725, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8598726114649682, | |
| "grad_norm": 0.4135420620441437, | |
| "learning_rate": 0.00015636518382138107, | |
| "loss": 0.0646, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8757961783439491, | |
| "grad_norm": 0.4710856080055237, | |
| "learning_rate": 0.00015461117736949577, | |
| "loss": 0.0538, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 0.428799033164978, | |
| "learning_rate": 0.00015283289404948976, | |
| "loss": 0.0593, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9076433121019108, | |
| "grad_norm": 0.4142407774925232, | |
| "learning_rate": 0.0001510311243799295, | |
| "loss": 0.0605, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9235668789808917, | |
| "grad_norm": 0.46757104992866516, | |
| "learning_rate": 0.0001492066693200096, | |
| "loss": 0.0634, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9394904458598726, | |
| "grad_norm": 0.4138467609882355, | |
| "learning_rate": 0.0001473603399134948, | |
| "loss": 0.0586, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 0.43409624695777893, | |
| "learning_rate": 0.00014549295692817778, | |
| "loss": 0.0579, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9713375796178344, | |
| "grad_norm": 0.44268661737442017, | |
| "learning_rate": 0.0001436053504910151, | |
| "loss": 0.0637, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9872611464968153, | |
| "grad_norm": 0.3840467929840088, | |
| "learning_rate": 0.00014169835971910238, | |
| "loss": 0.0511, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0031847133757963, | |
| "grad_norm": 0.4400401711463928, | |
| "learning_rate": 0.00013977283234665273, | |
| "loss": 0.0477, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.019108280254777, | |
| "grad_norm": 0.3332836925983429, | |
| "learning_rate": 0.00013782962434814492, | |
| "loss": 0.0478, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.035031847133758, | |
| "grad_norm": 0.3282802700996399, | |
| "learning_rate": 0.00013586959955780824, | |
| "loss": 0.0508, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0509554140127388, | |
| "grad_norm": 0.6200388073921204, | |
| "learning_rate": 0.00013389362928561317, | |
| "loss": 0.0479, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0668789808917198, | |
| "grad_norm": 0.5026119947433472, | |
| "learning_rate": 0.0001319025919299394, | |
| "loss": 0.0402, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0828025477707006, | |
| "grad_norm": 0.440782755613327, | |
| "learning_rate": 0.00012989737258709203, | |
| "loss": 0.0508, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0987261146496816, | |
| "grad_norm": 0.34029659628868103, | |
| "learning_rate": 0.0001278788626578407, | |
| "loss": 0.044, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 0.46119582653045654, | |
| "learning_rate": 0.00012584795945115603, | |
| "loss": 0.0426, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1305732484076434, | |
| "grad_norm": 0.47696712613105774, | |
| "learning_rate": 0.0001238055657853198, | |
| "loss": 0.0515, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 0.49719110131263733, | |
| "learning_rate": 0.00012175258958658564, | |
| "loss": 0.0453, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1624203821656052, | |
| "grad_norm": 0.36977869272232056, | |
| "learning_rate": 0.0001196899434855693, | |
| "loss": 0.0529, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.178343949044586, | |
| "grad_norm": 0.3333928883075714, | |
| "learning_rate": 0.00011761854441154767, | |
| "loss": 0.0429, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.194267515923567, | |
| "grad_norm": 0.3594333231449127, | |
| "learning_rate": 0.0001155393131848467, | |
| "loss": 0.044, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2101910828025477, | |
| "grad_norm": 0.33035799860954285, | |
| "learning_rate": 0.00011345317410749964, | |
| "loss": 0.0425, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.2261146496815287, | |
| "grad_norm": 0.431429386138916, | |
| "learning_rate": 0.00011136105455235766, | |
| "loss": 0.0395, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2420382165605095, | |
| "grad_norm": 0.34141987562179565, | |
| "learning_rate": 0.00010926388455083522, | |
| "loss": 0.0455, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2579617834394905, | |
| "grad_norm": 0.48105695843696594, | |
| "learning_rate": 0.00010716259637947357, | |
| "loss": 0.033, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 0.40072181820869446, | |
| "learning_rate": 0.0001050581241455064, | |
| "loss": 0.0404, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2898089171974523, | |
| "grad_norm": 0.5137969851493835, | |
| "learning_rate": 0.00010295140337161146, | |
| "loss": 0.0388, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.305732484076433, | |
| "grad_norm": 0.6727393865585327, | |
| "learning_rate": 0.00010084337058003303, | |
| "loss": 0.0455, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.321656050955414, | |
| "grad_norm": 0.36769551038742065, | |
| "learning_rate": 9.873496287626019e-05, | |
| "loss": 0.0416, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.3375796178343948, | |
| "grad_norm": 0.40487590432167053, | |
| "learning_rate": 9.662711753244551e-05, | |
| "loss": 0.0394, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3535031847133758, | |
| "grad_norm": 0.3262194097042084, | |
| "learning_rate": 9.452077157074994e-05, | |
| "loss": 0.0371, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3694267515923566, | |
| "grad_norm": 0.49285051226615906, | |
| "learning_rate": 9.241686134679867e-05, | |
| "loss": 0.0434, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3853503184713376, | |
| "grad_norm": 0.4286457896232605, | |
| "learning_rate": 9.031632213343339e-05, | |
| "loss": 0.0388, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.4012738853503186, | |
| "grad_norm": 0.4375673234462738, | |
| "learning_rate": 8.822008770494572e-05, | |
| "loss": 0.0419, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.4171974522292994, | |
| "grad_norm": 0.34244051575660706, | |
| "learning_rate": 8.612908992197705e-05, | |
| "loss": 0.0478, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 0.325065016746521, | |
| "learning_rate": 8.404425831726894e-05, | |
| "loss": 0.04, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4490445859872612, | |
| "grad_norm": 0.4267027676105499, | |
| "learning_rate": 8.196651968244826e-05, | |
| "loss": 0.0377, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4649681528662422, | |
| "grad_norm": 0.5654813051223755, | |
| "learning_rate": 7.989679765603108e-05, | |
| "loss": 0.0445, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.480891719745223, | |
| "grad_norm": 0.39815276861190796, | |
| "learning_rate": 7.783601231282812e-05, | |
| "loss": 0.0553, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.4968152866242037, | |
| "grad_norm": 0.46311256289482117, | |
| "learning_rate": 7.578507975493448e-05, | |
| "loss": 0.0373, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.5127388535031847, | |
| "grad_norm": 0.37769854068756104, | |
| "learning_rate": 7.374491170448525e-05, | |
| "loss": 0.0368, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.5286624203821657, | |
| "grad_norm": 0.2921994924545288, | |
| "learning_rate": 7.17164150983584e-05, | |
| "loss": 0.0428, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5445859872611465, | |
| "grad_norm": 0.48072826862335205, | |
| "learning_rate": 6.970049168500474e-05, | |
| "loss": 0.0329, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5605095541401273, | |
| "grad_norm": 0.37165966629981995, | |
| "learning_rate": 6.769803762358443e-05, | |
| "loss": 0.0376, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5764331210191083, | |
| "grad_norm": 0.309915155172348, | |
| "learning_rate": 6.570994308558812e-05, | |
| "loss": 0.032, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 0.41205787658691406, | |
| "learning_rate": 6.373709185911998e-05, | |
| "loss": 0.036, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.60828025477707, | |
| "grad_norm": 0.45491883158683777, | |
| "learning_rate": 6.17803609560181e-05, | |
| "loss": 0.0453, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.6242038216560508, | |
| "grad_norm": 0.4330281615257263, | |
| "learning_rate": 5.98406202219875e-05, | |
| "loss": 0.0319, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6401273885350318, | |
| "grad_norm": 0.2400377243757248, | |
| "learning_rate": 5.791873194991872e-05, | |
| "loss": 0.0342, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6560509554140128, | |
| "grad_norm": 0.26502475142478943, | |
| "learning_rate": 5.601555049656382e-05, | |
| "loss": 0.0318, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6719745222929936, | |
| "grad_norm": 0.2682415246963501, | |
| "learning_rate": 5.41319219027404e-05, | |
| "loss": 0.0301, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6878980891719744, | |
| "grad_norm": 0.38560014963150024, | |
| "learning_rate": 5.226868351723244e-05, | |
| "loss": 0.0432, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.7038216560509554, | |
| "grad_norm": 0.1889456957578659, | |
| "learning_rate": 5.042666362455498e-05, | |
| "loss": 0.0255, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.7197452229299364, | |
| "grad_norm": 0.3188437223434448, | |
| "learning_rate": 4.860668107674823e-05, | |
| "loss": 0.037, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.7356687898089171, | |
| "grad_norm": 0.7025167942047119, | |
| "learning_rate": 4.6809544929365004e-05, | |
| "loss": 0.0357, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 0.39571383595466614, | |
| "learning_rate": 4.503605408181286e-05, | |
| "loss": 0.0273, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.767515923566879, | |
| "grad_norm": 0.2839336693286896, | |
| "learning_rate": 4.3286996922211034e-05, | |
| "loss": 0.0302, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.78343949044586, | |
| "grad_norm": 0.3293486535549164, | |
| "learning_rate": 4.156315097692037e-05, | |
| "loss": 0.0308, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.799363057324841, | |
| "grad_norm": 0.3942393958568573, | |
| "learning_rate": 3.986528256490141e-05, | |
| "loss": 0.0293, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.8152866242038217, | |
| "grad_norm": 0.31392380595207214, | |
| "learning_rate": 3.8194146457054655e-05, | |
| "loss": 0.0249, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8312101910828025, | |
| "grad_norm": 0.3941175639629364, | |
| "learning_rate": 3.655048554069478e-05, | |
| "loss": 0.032, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.8471337579617835, | |
| "grad_norm": 0.2796083092689514, | |
| "learning_rate": 3.4935030489306883e-05, | |
| "loss": 0.034, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8630573248407645, | |
| "grad_norm": 0.30333149433135986, | |
| "learning_rate": 3.334849943773323e-05, | |
| "loss": 0.0351, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8789808917197452, | |
| "grad_norm": 0.3254244029521942, | |
| "learning_rate": 3.179159766293282e-05, | |
| "loss": 0.0298, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.894904458598726, | |
| "grad_norm": 0.20831990242004395, | |
| "learning_rate": 3.0265017270457775e-05, | |
| "loss": 0.0288, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 0.31671157479286194, | |
| "learning_rate": 2.8769436886784408e-05, | |
| "loss": 0.0334, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.926751592356688, | |
| "grad_norm": 0.5833495259284973, | |
| "learning_rate": 2.730552135763632e-05, | |
| "loss": 0.0425, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.9426751592356688, | |
| "grad_norm": 0.32988494634628296, | |
| "learning_rate": 2.5873921452433915e-05, | |
| "loss": 0.0328, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.9585987261146496, | |
| "grad_norm": 0.20500314235687256, | |
| "learning_rate": 2.4475273575000936e-05, | |
| "loss": 0.0273, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9745222929936306, | |
| "grad_norm": 0.33419910073280334, | |
| "learning_rate": 2.3110199480657525e-05, | |
| "loss": 0.0264, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9904458598726116, | |
| "grad_norm": 0.20512598752975464, | |
| "learning_rate": 2.1779305999824884e-05, | |
| "loss": 0.0287, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.0063694267515926, | |
| "grad_norm": 0.4389275312423706, | |
| "learning_rate": 2.0483184768264596e-05, | |
| "loss": 0.0306, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.022292993630573, | |
| "grad_norm": 0.18635383248329163, | |
| "learning_rate": 1.9222411964072884e-05, | |
| "loss": 0.028, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.038216560509554, | |
| "grad_norm": 0.32153043150901794, | |
| "learning_rate": 1.799754805154603e-05, | |
| "loss": 0.0274, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.054140127388535, | |
| "grad_norm": 0.448691189289093, | |
| "learning_rate": 1.6809137532031704e-05, | |
| "loss": 0.0306, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 0.38386714458465576, | |
| "learning_rate": 1.565770870187585e-05, | |
| "loss": 0.0247, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0859872611464967, | |
| "grad_norm": 0.18705415725708008, | |
| "learning_rate": 1.4543773417573925e-05, | |
| "loss": 0.025, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.1019108280254777, | |
| "grad_norm": 0.23585732281208038, | |
| "learning_rate": 1.3467826868229994e-05, | |
| "loss": 0.0265, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.1178343949044587, | |
| "grad_norm": 0.36259782314300537, | |
| "learning_rate": 1.243034735542512e-05, | |
| "loss": 0.0305, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.1337579617834397, | |
| "grad_norm": 0.34997355937957764, | |
| "learning_rate": 1.1431796080593283e-05, | |
| "loss": 0.032, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.1496815286624202, | |
| "grad_norm": 0.3579126298427582, | |
| "learning_rate": 1.0472616939998492e-05, | |
| "loss": 0.024, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.1656050955414012, | |
| "grad_norm": 0.41504374146461487, | |
| "learning_rate": 9.553236327405246e-06, | |
| "loss": 0.0259, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1815286624203822, | |
| "grad_norm": 0.2632668912410736, | |
| "learning_rate": 8.674062944529216e-06, | |
| "loss": 0.0301, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1974522292993632, | |
| "grad_norm": 0.3617275059223175, | |
| "learning_rate": 7.835487619352811e-06, | |
| "loss": 0.0232, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.213375796178344, | |
| "grad_norm": 0.3129204213619232, | |
| "learning_rate": 7.037883132386547e-06, | |
| "loss": 0.0355, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 0.2824781835079193, | |
| "learning_rate": 6.2816040509530165e-06, | |
| "loss": 0.0269, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.245222929936306, | |
| "grad_norm": 0.20426948368549347, | |
| "learning_rate": 5.566986571567401e-06, | |
| "loss": 0.0355, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.261146496815287, | |
| "grad_norm": 0.22026273608207703, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.0267, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.2770700636942673, | |
| "grad_norm": 0.27001041173934937, | |
| "learning_rate": 4.263988462479484e-06, | |
| "loss": 0.0286, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2929936305732483, | |
| "grad_norm": 0.22612400352954865, | |
| "learning_rate": 3.676187067922421e-06, | |
| "loss": 0.0263, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.3089171974522293, | |
| "grad_norm": 0.3205125033855438, | |
| "learning_rate": 3.131205488210409e-06, | |
| "loss": 0.0311, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.3248407643312103, | |
| "grad_norm": 0.2743224501609802, | |
| "learning_rate": 2.6292859896079213e-06, | |
| "loss": 0.0233, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.340764331210191, | |
| "grad_norm": 0.23394609987735748, | |
| "learning_rate": 2.170651695549786e-06, | |
| "loss": 0.0161, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.356687898089172, | |
| "grad_norm": 0.23162613809108734, | |
| "learning_rate": 1.7555064874538397e-06, | |
| "loss": 0.0243, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.372611464968153, | |
| "grad_norm": 0.3181140720844269, | |
| "learning_rate": 1.3840349140874619e-06, | |
| "loss": 0.0224, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 0.29145529866218567, | |
| "learning_rate": 1.0564021095281652e-06, | |
| "loss": 0.0345, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.404458598726115, | |
| "grad_norm": 0.321485698223114, | |
| "learning_rate": 7.727537197548707e-07, | |
| "loss": 0.0205, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.4203821656050954, | |
| "grad_norm": 0.2429160177707672, | |
| "learning_rate": 5.332158379024122e-07, | |
| "loss": 0.0217, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.4363057324840764, | |
| "grad_norm": 0.22818560898303986, | |
| "learning_rate": 3.3789494820803957e-07, | |
| "loss": 0.0219, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.4522292993630574, | |
| "grad_norm": 0.2582818865776062, | |
| "learning_rate": 1.8687787867489592e-07, | |
| "loss": 0.0186, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.468152866242038, | |
| "grad_norm": 0.2430548369884491, | |
| "learning_rate": 8.023176247348163e-08, | |
| "loss": 0.0344, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.484076433121019, | |
| "grad_norm": 0.38517120480537415, | |
| "learning_rate": 1.8004008098226887e-08, | |
| "loss": 0.0225, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.498407643312102, | |
| "step": 1569, | |
| "total_flos": 5.687859742190496e+16, | |
| "train_loss": 0.06483107545027846, | |
| "train_runtime": 706.8123, | |
| "train_samples_per_second": 35.517, | |
| "train_steps_per_second": 2.22 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1569, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.687859742190496e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |